As many as one in three older adults may experience insomnia. Older adults are the most frequent users of hypnotic medications. Although safer, use of even the latest sleeping pills can lead to cognitive impairment and risk of falls. Thus, it isnot surprising that non-pharmacological treatments for insomnia have been pursued as alternatives to medications, with some suggesting they should be the first line of therapy. We propose a randomized clinical trial to evaluate the relative efficacy and effectiveness of specific components of CBT-I: sleep restriction (SR) and cognitive therapy (CT) in comparison to combined SR and CT (SR+CT). We hypothesize that because the proposed mechanisms of action of CT versus SR substantially differ, their combination may have additive effects. Even though the mechanisms of action of SR and CT may differ, no data exists to document that the addition of one to the other provides more overall clinical benefit than either intervention alone. Finally, to better understand how and in whom SR and CT work, we plan to formally evaluate selected mediators and moderators of the clinical effect including cognitive and physiological measures of anxiety and arousal. Three treatments (SR, CT, and SR+CT) will be compared in a randomized clinical trial with a parallel groups design. Efficacy and effectiveness data will be collected prior to the beginning of treatment, after six weeks of treatment, and at the end of a six-month follow-up. This five-year application will ultimately include 150 subjects (three groups of 50 each) who are at least 60 years of age. Data will include a sleep log, polysomnography, mediator and moderator measures, and quality of life measures on all subjects in each group. These efforts follow the National Institute of Mental Health Strategic Plan Strategy 3.1 to develop innovative interventions and designs for intervention studies, in this case, to promote a new intervention trial that focuses on the mediators and moderators of treatment response.